To date, SFRT has been demonstrated to have promising clinical effects such as a response rate over 90% and complete response (CR) rate of 27% when used alone,12 which highlights the positive impact of SFRT on partial tumors. Furthermore, a previous study reported abscopal effects and bystander effects when the hypoxic core of tumor was targeted by SBRT level of radiation 16.
Based on this evidence, we hypothesize that the larger is the volume of central portion of tumor receiving an ablative dose, the larger probability that the radiation treatment would trigger biological CKBE effects. Concurrently, maintaining low-dose fractionated radiation to the tumor edge can help to avoid most of the tissue toxicity seen at ablating the whole tumor.
Based on these considerations, in this study, we propose a new treatment methodology called Stereotactic Core Ablative Radiation Therapy, or SCART, for bulky primary or metastatic tumors. The proposed methodology is based on the principles of SFRT and SBRT-based methods are used to deliver an ablative radiation dose to one large central portion of target (called SCART-TV or STV) while keeping the dose to surrounding normal tissue to a relatively low level. In prioritizing tissue safety, we hope to generate a large portion within the target that is allowed by the modern technology. SCART designs an ultra-heterogeneous dose distribution which transits from a high, ablative SCART dose in STV to a modest dose in TTV then to a low, tissue-safe dose outside of GTV border. Cancer cells in the STV tend to be hypoxic and cold in immune microenvironment. Killing these cells requires the ablative dose, and the radiation ablation is more likely to release anti-tumor antigens and trigger the bystander and abscopal effects10. Peripheral part of GTV surrounding the core STV is defined as TTV (Transitional Treatment Volume) and the TTV essentially serves as a “cushion” between ablated target and surrounding tissue. Cancer cells in the peripheral TTV tend to be normoxic and warm in immune micro-environment, while a modest dose has the potential to upregulate the immune micro-environment 17. This ultra-heterogeneous dose distribution serves multiple purposes.
[Figure 2 about here]
Figure 2: The concept of SCART. GTV (Gross Tumor Volume) is the entire bulky tumor volume (in blue). STV (SCART- Treatment Volume) is the central core segment of GTV receiving ablative SCART dose (in orange). TTV (Transitional Treatment Volume) is the rest of GTV volume outside of STV and is modestly irradiated; its ultra-heterogeneous dose transits from a high, ablative SCART dose in STV to a modest dose in TTV then to a low, tissue-safe dose outside of GTV border
The three key aspects of SCART are as follows:
1) Tissue Safety. Considering that most patients with bulky tumors tend to be fragile, and the tissues surrounding bulky tumors already underwent numerous treatments or are stretched, it is essential keep the dose at the GTV border low to avoid tissue damage. Accordingly, we limited the dose at 5Gy per fraction and identify 1 ~ 3 fractions, which is below the tolerance of most of the surrounding normal tissue. In addition, there may be a serial organ (such as a vessel or the trachea) residing at the center of GTV that should not be overlapped with TTV.
2) Ablative dose. The goal of SCART is not to treat the whole GTV based on the theory of DNA radiation damage, but to irradiate its central part to trigger Cancer Killing Biological Effects (CKBE). According to available reports, such responses can be triggered more likly by an ablative dose (> 8Gy per fraction), while the conventional fractionated 2Gy per fraction is incapable of doing so11. We typically put the SCART dose at 15 ~ 24Gy x 3 fractions.
3) Large volume of the ablated STV. We expected that a large volume of STV would not only cause a direct DNA-damage of large amount of cancer cells, but would also release a large amount of tumor-related antigen (tumor vaccine), thereby more likely triggering anti-tumor immune response. We typically limited the GTV border dose to 3 or 5Gy per fraction and planned for an ablation dose of 15 ~ 24Gy per fraction at STV. We designed a process, as explained in the following section, to generate a plan with a fast dose fall-off at TTV, which would, in turn, allow for a larger volume of STV.
Comparison of various SFRT technologies: GRID, Lattice, and SCART
A clinical example of comparing three SFRT methods – GRID, Lattice, and SCART – is shown in Fig. 3. The parameters comparing the three methods are summarized in Table 1. The results revealed that the volume of tumor that received the ablative prescription dose in SCART (67.6 cc) is 10–100 times larger than the corresponding volumes in GRID (2.26 cc) and Lattice (0.12 cc) methodologies.
[Figure 3 about here]
Table 1
summerizes the three SFRT methods on a clinical example of a 672 cc endometrial sarcoma with 15Gy x 3 as the prescription dose.
Test Case Comparison
|
Grid
|
Lattice
|
SCART
|
Dmax (cGy)
|
16.5 Gyx3
|
15Gy x3
|
17.3 Gyx3
|
MU
|
3100 x3
|
4469 x3
|
3398 x3
|
V100% (cc)
|
2.26
|
0.12
|
67.6
|
[Table 1 about here]
The implementation of SCART
SCART treatment can be implemented by modern machines such as a linear accelerator (Linac), a robotic radiosurgery system (Cyberknife), or a particle accelerator (Proton). Yet the primary technology used in the present study is Volumetric Modulated Arc Therapy (VMAT), which is most commonly applied in SCART with the Linac rotating around the patient who receives continuous doses of radiation towards a tumor (STV in SCART).
Typical SCART prescription dose is 15 ~ 24Gy x3, while the surrounding tissue dose tolerance is set at 5Gy x 3. In the present study, the patients underwent CT-based simulation with 1 ~ 2 mm thin slice CT for treatment planning. Contrasted CT, MRI, and PET/CT are often acquired and fused with sim CT to help tumor and organ at risk (OAR) target delineation and contouring. In these images, gross target volume (GTV) is defined as bulky visible disease. Clinical target volume (CTV) is assumed to be identical to GTV and is not highlighted in this study. STV is defined as the central region inside of GTV. In our process, we define a STV as the ablation target first, then optimize the VMAT plan with the tissue sparing as top priority without emphasizing the conformality of STV coverage.
Design STV
The STV is predefined prior the VMAT plan optimization as the real target of ablation. In our planning, the maximum opening of radiation MLC was limited to the size of STV and these radiation fields pass through TTV and (only) converged at STV. This arrangement is expected to generate the high dose gradient in the region right outside of STV, defined as the TTV sitting between STV and GTV border. (see Fig. 4).
Figure 4a. Illustration of Linac (VMAT) irradiates the STV (in orange) from a full 360 degree arc. The maximum radiation field sizes from various angles are limited to its projection of STV. These fields pass through TTV and intersect at STV. The dose intensity falls off quickly at TTV to a modest dose at the edge of GTV, which is safe to the surrounding tissue. 4.b Dose cloud of a hypothetical elliptical GTV and an elliptical STV at the core of GTV in axial view, along with the spindle shape of SST in SI direction (Sagittal and Coronal view)
[Figure 4 about here]
The STV usually locates at the center of GTV as previously discussed. Occasionally, we relocate the STV slightly to better accommodate tissue sparing.
The shape is similar to GTV in axial view and it looks like a spindle in SI view in VMAT planning.
The optimal dimension of STV is not instinctively straightforward to define. Dose gradient at TTV is related to the size of STV. With an increase of STV, the radiation field size is larger, and the dose falls off slower at TTV. Meanwhile, when the STV is larger, TTV is smaller, allowing less space for the ablative dose at STV border to fall off to a low, tissue-safe constraint dose at GTV border. Our intention in the present study was to find the optimal dimension of STV so that the ablative SCART prescription dose at border of STV would precisely fall off to the tissue constraints dose at the border of GTV. In that case, the STV would have the largest possible volume.
The methodology to determinate the dimension of STV (in axial view) is described in supplementary. The supplementary also includes two interesting findings which make the determination relatively easy based on a simply equation (Eq. 1)
- STV/GTV pretty much independent with dimension of GTV
- STV/GTV depends on the tissue dose constraints / SCART prescription dose. As a rule of thumb for VMAT-based SCART planning, in axial view, we used Eq. (1):
STV dimension = GTV dimension * Tissue Dose Constraints / Prescription Dose (1)
In SI direction, STV was slightly shorter than GTV, and the STV typically looked like a spindle in the sagittal and coronal views.
After designing and correcting/editing STV, we ensured that there was no overlap with the serial organs, such as vessels or the trachea.
Treatment planning, beam arrangement and treatment delivery
One or two co-planar 360 degree 6 MV photon arcs were used with the VMAT technology. Heterogeneity corrections were used in the definitions of all doses. To aid the optimization process in minimizing the dose to critical structures, non-anatomical dose constraining ring structures were incorporated. The ring structure and tissue constraints were given higher priority and the STV coverage was given second priority, in the optimization. Caution was given to further spare important organ if previously heavily treated.
The SCART was delivered every other day until the intended dose level was achieved or discontinuation due to toxicity. Cone beam (CBCT) image guidance was acquired in every treatment. All patients in our Phase 1 trial completed the treatment with average beam-on time of 8.9min and average treatment time of 18.5min.
SCART vs. the conventional Boost radiation
SCART pursues the ablation of the center (cold immune micro-environment) and spares the peripheral areas of the tumor (active and warm immune micro-environment) that are only irradiated with a heterogeneous, relatively lower, non-ablative dose. By contrast, the conventional central boost radiation strategy irradiates the whole tumor (mostly non-ablative) and gives extra dose to the center in order to increase the chance of tumor control. Technically, each SCART radiation field is narrow and only directly irradiates the core part of tumor (STV), but not the entire tumor. The dose out of STV falls off fast and the dose at STV is significantly (3 ~ 8 fold) higher than that at the border of tumor. By contrast, the boost strategy unevenly irradiates the entire tumor and gives certain extra dose to the center, which is modestly (10 ~ 100%) higher than the dose at the border of tumor (see Fig. 5).
[Figure 5 about here]
Clinical Case example
A 72-y old male COPD patient with a bulky squamous cell carcinoma in the left low lobe (8 x 8 x 13 cm, 399 cc) underwent multiple courses of systematic therapy and was in pain. The patient was not a surgical candidate and refused the conventional 30 fractions of the radiation therapy. He received 24Gy x3fractions SCART treatments, after which, at 3 months post-treatment, the tumor shrank to 4 x 6 x 11cm, 133cc and the patients’ pain was significantly relieved. This patient subsequently received SBRT of 5Gy x8 to the residue disease, with further control of the disease (see Fig. 6).
[Figure 6 about here]
Clinical Outcome of Phase I Clinical Trial
We performed a prospective, multi-center, dose-escalation phase 1 IRB approved trial of SCART with and without additional External Beam Radiation Therapy (EBRT) for bulky metastatic or recurrent cancer. A total of 19 patients aged over 18 years old with 21 biopsy-confirmed recurrent or metastatic bulky tumors were included in the sample. All patients had measurable disease documented by CT and/or PET amenable for SCART radiation with the shortest axis of 3 cm or longer.
The average age of the patients was 66 years and 58.8% were male. Half of the sample had a tumor either in the liver (25.0%) or in the lung (25.0%). Almost a quarter (23.5%) of the patients had adenocarcinoma, and most tumors (70.0%) were Stage IV. The average tumor size decreased from 301 cm3 to 118 cm3 (71% shrinkage) with the average follow-up of 15m (2 ~ 36). Regarding the overall survival shown in Fig. 7, of 19 patients, twelve died (63%). The proportion of patients free from death at one year after the beginning of SCART treatment was 50%, and 40% at two years after. The median time to death was 15 months. Long follow-up showed that 14/21 tumors achieved PR, 2/21 CR, 3/21 SD, and 1/21 PD, leading to an encouraging local control of 95%. Out of 19 patients, eight cases were grade I toxicity and one was grade II toxicity; none cases of level III toxicity were found.
[Figure 7 about here]